Development of Machinery Survey Requirements Based on Reliability

ABS TECHNICAL PAPERS 2005

Development of Machinery Survey Requirements Based on Reliability-Centered Maintenance Robert M. Conachey, Member, American Bureau of Shipping

Originally presented at the 2005 SNAME Marine Technology Conference & Expo. Reprinted with the permission of the

Society of Naval Architects and Marine Engineers (SNAME). Material originally appearing in SNAME publications cannot be reprinted without written permission from the Society,

601 Pavonia Ave., Jersey City, NJ 07306.

ABSTRACT

This paper discusses the technical background of the American Bureau of Shipping (ABS) “Guide for Survey Based on Reliability-centered Maintenance” to improve reliability for vessels’ machinery systems and receive credit towards certain machinery survey requirements. Risk assessment techniques and RCM analysis are used to provide a process to optimize maintenance tasks and achieve optimal reliability. A process for spare holding requirements incorporating risk is discussed. A sustainment process was developed so the operator can keep the preventative maintenance tasks current as the system ages, new failure modes or system modifications occur. The approaches taken to address the lack of quantitative data for equipment failures, consistency of analyses among operators and types of consequences and descriptions related to their severity are discussed.

NOMENCLATURE Failure modes - The failure mode describes how equipment can fail and potentially result in a functional failure. Failure mode can be described in terms of an equipment failure cause (e.g., pump bearing seizes), but is typically described in terms of an observed effect of the equipment failure (e.g., pump fails off). Failure modes, effects and criticality analysis (FMECA) – FMECA is an inductive reasoning approach that considers how the failure mode of each system component can result in system performance problems. It is expanded to include failure frequencies so failure modes may be ranked according to risk. Functional failures - A functional failure is a description of how the equipment is unable to perform a specific function to a desired level of performance. Risk - Risk is composed of two elements, frequency and consequence. Risk is defined as the product of the frequency with which an event is anticipated to occur and the severity of the consequence of the event’s outcome.

INTRODUCTION By applying Reliability-centered maintenance

(RCM) principles, maintenance is evaluated and applied

in a rational manner that provides the most value to a vessel’s owner/operator. Accordingly, improved equipment and system reliability on board vessels and other marine structures can be expected by applying this philosophy.

RCM is also a part of overall risk management so that the risk of undesirable end events associated with equipment failures can be effectively managed by the maintenance program. This failure management is achieved by allocating maintenance resources to equipment maintenance according to risk impact on the vessel. For example, RCM analysis can be employed to:

• Identify functional failures with the highest risk, which will then be focused on for further analyses;

• Identify equipment items and their failure modes that will cause high-risk functional failures; and

• Determine a maintenance strategy that will reduce risk to acceptable levels.

A brief overview of the RCM process as applied in the Guide for Survey Based on Reliability-centered Maintenance (RCM Guide) by the American Bureau of Shipping (ABS) is provided (ABS 2003). The RCM Guide lists the requirements for the ABS RCM Program, a voluntary Program that enables vessel operators to receive credit towards certain machinery survey requirements in order to maintain a vessel’s

Development of Machinery Survey Requirements Based on Reliability-Centered Maintenance 229


classification. A companion document to the RCM Guide, Guidance Notes on Reliability-centered Maintenance was published to provide additional information related to maintenance and risk analysis (ABS 2004). This paper discusses the ABS approach taken to apply the principles of this maintenance philosophy.

OVERVIEW OF RCM PRINCIPLES RCM is a process of systematically analyzing an

engineered system to understand: • system functions and impact of functional

failures • equipment failure modes and causes that can

result in functional failures • optimal strategy for managing potential

failures, including maintenance to prevent the failures from occurring or to detect potential failures before a failure occurs, and

• spares holding requirements. ABS requires the following analytical tools to be employed when performing the RCM analyses:

• Failure modes, effects, and criticality analysis (FMECA),

• RCM task selection flow diagram, • Risk-based decision making tools (e.g., risk

matrix). In addition, the following system expertise is needed to successfully and efficiently perform the analysis:

• Design, engineering, and operational knowledge of the system,

• Condition-monitoring techniques, planned maintenance actions, failure finding techniques,

• Other proactive maintenance practices (e.g., lubrication).

Equipment failure basics Since 1978, ABS has cooperated with owners/operators on developing and implementing preventative maintenance programs. The Bureau recognized an effective program improved machinery reliability. ABS issued its first Guide for Survey Based on Preventative Maintenance Techniques (PM Guide) in 1984 that listed the requirements for a preventative maintenance program and provided credit towards a vessel’s Special Periodical Machinery Survey for equipment enrolled in the program. Unlike the PM Guide, the RCM Guide provides owner/operators a process to create an effective preventative maintenance program applying risk principles and a maintenance task methodology. The RCM analysis process uses these tools and expertise to help establish the cause effect relationship between equipment failures and system performance (e.g., the FMECA) and then determine an effective failure management strategy (e.g., RCM task selection). A combination of one or more equipment failures and/or human errors causes a loss of system function.

Specifically, one of the focuses of reliability improvement is to manage the equipment failures that impact system performance (e.g., losses of system function). Therefore, an understanding of the factors that influence equipment failures is needed. The following factors usually influence equipment failure:

• Design error • Faulty material • Improper fabrication and construction • Improper operation • Inadequate maintenance • Maintenance errors

Therefore, maintenance is merely one of the many approaches to improving equipment reliability and hence system reliability. RCM analyses focus in reducing failures resulting from inadequate maintenance. In addition during the RCM analysis process, some equipment failures may be identified as the result of maintenance errors. In these cases the results of RCM analyses may suggest improvements for specific maintenance activities, such as improving the manner in which the maintenance procedures are carried out, improving worker performance through additional training or required skill level, or adding quality assurance/quality control tasks during the maintenance procedure to verify correct performance of critical maintenance tasks. Furthermore, RCM analyses may recommend design changes and/or operational improvements when equipment reliability cannot be ensured through maintenance.

Equipment failure rate and patterns One of the key concepts of RCM is that all equipment failures are not the same; therefore, the maintenance tasks necessary to prevent failures may require different strategies in order to successfully manage them. In fact, depending on the dominant system failure mechanisms, system operation, system operating environment, and system maintenance, specific equipment failure modes exhibit a variety of failure rates and patterns. First, let’s discuss the failure rate. The conditional probability failure rate or lambda (λ) is the probability that a failure occurs during the next instant of time given that the failure has not already occurred before that time. The conditional failure rate, therefore, provides additional information about the survival life and is used to illustrate failure patterns. For most equipment failure modes, the specific failure patterns are not known and fortunately detailed knowledge is not needed to make maintenance decisions. Nevertheless, certain failure characteristic information is needed to make maintenance decisions. These characteristics are:

• Wear-in failure – dominated by “weak” members related to problems such as manufacturing defects and installation/maintenance/startup errors. Also known as “burn in” or “infant mortality” failures.

230 Development of Machinery Survey Requirements Based on Reliability-Centered Maintenance


• Random failure – dominated by chance failures caused by sudden stresses, extreme conditions, random human errors, etc. (e.g., failure is not predictable by time) during the “useful life” of the equipment.

• Wear-out failure – dominated by end-of-useful life issues for equipment.

These failure characteristics are best illustrated by the failure pattern identified in Figure 1. By simply identifying which of the three equipment failure characteristics is representative of the equipment failure mode, one gains insight into the proper maintenance strategy. Understanding that equipment failure modes can exhibit different failure patterns has important implications when determining appropriate maintenance strategies. The literature has indicated there are six different failure patterns as shown in Table 1 (Nowlan/Heap 1978, Moubray 1997, and Smith 1993). We have listed the failure characteristic(s) too along with some representative examples:

• Pattern A – Bathtub Curve – Wear-In, Random Failure, Wear-Out

• Pattern B – Traditional Wear-Out – Random Failure, Wear Out

• Pattern C – Gradual Rise with no Distinctive Wear-out Zone - Random Failure

• Pattern D – Initial Increase with a Leveling Off – Random Failure

• Pattern E – Random – Random Failure • Pattern F – Infant Mortality- Wear-In, Random

Failure Those patterns that do not have distinctive wear-out regions (e.g., patterns C through F) may not benefit from maintenance tasks of rebuilding or replacing equipment items. There may actually be an increase in failures as a result of infant mortality (pattern F) and/or

human errors during maintenance tasks. If an equipment failure mode exhibits a wear-out pattern, rebuilding or replacing the equipment item may be an appropriate strategy. Finally, a basic understanding of failure rate helps in determining whether maintenance or equipment redesign is necessary and provides insight into frequency of maintenance tasks. Once one begins to understand how equipment fails and its failure rate and pattern, an understanding of maintenance task types and their relationship to the failure characteristics is needed.

Overview of Maintenance Task Types One of the primary objectives of the RCM analysis is to define a set of proactive maintenance tasks needed to manage potential equipment failures that can impact critical system performance. These tasks can manage these potential failures by:

• Detecting onset of failure with sufficient time to allow corrective action before the failure occurs, e.g. condition monitoring tasks,

• Preventing the failures before they occur, which are referred to in the RCM Program as planned maintenance tasks,

• Discovering and correcting hidden failures before they impact system performance, e.g. failure finding tasks.

• Applying operational restrictions or some other action, e.g. any applicable and effective task.

In addition, the RCM analysis might indicate the failure does not warrant any proactive maintenance and run-to-failure is acceptable. Also, RCM analyses should include routine servicing tasks to ensure the assumed failure rate and failure pattern are valid (e.g., failure rate and pattern for an un-lubricated bearing is drastically different from that of a lubricated bearing).

Figure 1: Equipment Life Periods

IIn fa n t

M o rta lity

IIU s e fu l

L ife

IIIW e a r o u t

T im eW e a r o u tB u rn in

Failu

reR

ate λ

(t)



Table 1: Six Classic Failure Rate Patterns

λ(t)

t

Pattern A – Bathtub: Infant mortality, then a constant or increasing failure rate, followed by a distinct wear-out zone Example: overhauled reciprocating engine

λ(t)

t

Pattern B – Traditional Wear-out: Constant or slowly increasing failure rate followed by a distinct wear-out zone Example: reciprocating engine, pump impeller

λ(t)

t

Pattern C – Gradual Rise with no Distinctive Wear-out Zone: Gradually increasing failure rate, but no distinct wear-out zone Example: gas turbine

λ(t)

t

Pattern D – Initial Increase with a Leveling off: Low failure rate initially, then a rapid increase to a constant failure probability Example: complex equipment under high stress with test runs after manufacture or restoration such as hydraulic systems

λ(t)

t

Pattern E – Random Failure: Constant failure rate in all operating periods Example: roller/ball bearings

λ(t)

t

Pattern F – Infant Mortality: High infant mortality followed by a constant or slowly rising failure rate Example: electronic components

RCM ANALYSIS PROCESS ABS reviewed the RCM analysis literature and concluded that what was available would need to be modified for marine applications. We also decided that the procedures selected would be in conformance with a recognized standard. We selected SAE JA 1011, Evaluation Criteria for Reliability-Centered Maintenance (RCM) Processes (Society of Automotive Engineers 1999). We believed evaluation of risk was necessary in order to rank the relative importance of the failures analyzed, so we incorporated risk in the failure modes and effects analysis adopting the approach in the IMO International Code for High Speed Craft (IMO 2001). We also felt it necessary to include a process to determine spare parts requirements using risk assessment. Any successful long-term maintenance program must include a feedback mechanism so we included requirements for a sustainment procedure. Accordingly, the basic steps of the RCM analysis process are:

1. Identify operating modes and corresponding operating context

2. Define vessel systems 3. Develop system block diagrams and identify

functions 4. Identify functional failures 5. Conduct failure modes, effects, and criticality

analysis (FMECA) 6. Select failure management tasks 7. Determine spare parts holdings 8. Develop RCM sustainment process 9. Document the analysis 10. Implement RCM Onboard

Step Nos. 1 through 4 – System Modeling, Functions and Functional Failures For this part of the analysis, how the vessel is to be operated (operating mode) and the manner in which the vessel’s machinery systems are operated (operating context) are determined. The vessel’s systems are



modeled as a hierarchy for the purposes of performing the FMECA. An example is shown in Figure 2. For consistency, ABS has named these hierarchy levels, in descending order as follows: functional group, system, sub-system, equipment item, and component. A component is defined as:

• the lowest level that can be identified for its contribution to the overall functions of the functional group,

• being identifiable for its failure modes, and • the most convenient physical unit that can be

considered for the preventative maintenance plan.

The system block diagrams serve as an aid to visualize the hierarchical structure and identify the various system functions as shown in the example in Figure 3. Then, the various functional failures associated with each function are identified. Fortunately, the arrangement of the components within many systems such as, fuel oil, cooling water, lubricating oil, is similar among vessels because of classification and statutory requirements. Accordingly, ABS has created “templates” for participants in the ABS RCM Program which are partially completed FMECAs for various systems subject to our Rules and Special Periodical Machinery Survey requirements on vessels. These systems are primarily associated with propulsion, directional control, vessel safety and cargo handling. The templates permit those performing the analysis to reduce the analysis time, and ensure consistency of analyses submitted to ABS.

Step No. 5 – Conduct FMECA For Step No. 5, ABS requires the application of a bottom-up FMECA. An example format is shown in Table 2. We selected the bottom-up format instead of the top-down format because during development of the preventative maintenance plan for each system component, there is less of a chance a component will be omitted. The top-down format is useful when designing new systems to determine the risk associated with various functional failures. If one chooses to apply the top-down format for existing systems, it is necessary to identify all system functions otherwise it is

likely a component may be omitted that contributes to that system function. To ensure consistency among the analyses received, in the developed templates we have listed suggested failure modes for each component. For some operating modes/contexts these failure modes may not be applicable and can be indicated so in the analysis. In some cases the failure modes listed may not have been considered by the analysis team. ABS has provided a list of suggested failure modes for ten groups of equipment and components in Appendix 2 of the RCM Guide.

ABS decided for consistency among analyses received to require the End Effect descriptions to be the effect on the functional group(s). A consolidated example format from the RCM Guide is shown in Table 3. The severity level is defined for at least four levels from no effect, two progressive functional degradations to complete loss of function. Four levels is the minimum to ensure meaningful risk ranking. An additional severity level or two may be considered but greater care is necessary in severity level definition. The traditional approach is to define severity levels based on an order of magnitude in economic terms (i.e. $10,000, $100,000, etc.). Some would consider the approach in the RCM Guide as determining the intermediate effect, not the end effect. However, as part of the certification process it is straightforward for determining failure effects on functional groups when a component failure occurs. Attempting to determine the ultimate end effects from a vessel’s complete loss of propulsion, such as grounding and considering other end effects such as pollution caused by rupture of the fuel oil storage tanks, loss of revenue, etc., is much more subjective and therefore difficult to evaluate. Such end effects would be dependent on the operating mode of the vessel, geographic location, etc. If desired, the owner/operator can extend the analysis to assess business risks. A brief example of this approach along with an estimate in risk reduction is provided in Appendix 1. However, the RCM Program does not require analyses addressing business risks be submitted because of the proprietary business information they will likely include.

TABLE 2: Example Bottom-up FMECA Worksheet

No.: XX Description: Pump

Item Failure Mode Causes Failure Characteristic

Local Effects Functional Failures

End Effects

No.: XX Description: Pump

Item Matrix Severity Current Likelihood Current Risk Failure Detection/ Corrective Measures



The other element of risk is the likelihood or frequency of the failure mode. Many efforts have been made and are currently underway to collect failure rate data for machinery. Obtaining quantitative failure data is problematic: published data is scant, reliability databases are available only to subscribing members of an industry, manner of data collection is unknown, failure modes are not identified, etc. ABS decided to take a qualitative approach by recommending frequency ranges as shown in Table 4. As the FMECA is developed, we believe the team members can estimate failure mode frequencies based on events occurring within their operating fleet or collective memory. We expect that with time as failure data are collected electronically in maintenance and repair software, quantitative data can be determined and compared with the estimated qualitative data in the initial analysis.

Step No. 6 – Select Failure Management Tasks There are several RCM task selection flow diagrams in the literature (Ministry of Defence (UK) 1999, Moubray 1997, Naval Air Systems Command (USA) 2001, Society of Automotive Engineers 1999). ABS considered all of them and adapted the appropriate features from them for application to the marine industry. The RCM Task Selection Flow Diagram is shown in Figure 4.

The ABS task selection process is similar to other selection processes with respect to requiring one-time changes for failure modes with the highest risk, and a run-to-failure strategy for failure modes with the lowest risk. For failure modes with risks between the extremes, maintenance task types in the following order are considered: condition monitoring, planned maintenance, combination condition monitoring and planned maintenance, any applicable and effective task, or one-time change. For hidden failure modes, failure finding tasks are specified. Unlike other published task selection flow processes we have included additional procedures as shown in the continuation for Figure 4. These include a procedure to specify a maintenance task(s) to address all causes associated with the failure mode for evaluation. The risk is re-evaluated for the selected maintenance tasks and any one-time changes associated with a failure mode. If the risk level meets the acceptance criteria, the next failure mode is evaluated. If not, the maintenance tasks and one-time changes are re-evaluated to seek a reduction in the risk to acceptable criteria. These criteria would include: a reduction in or at least the same level of risk compared to no maintenance or present maintenance tasks; the failure mode does not result in the highest risk occurring.



Figure 2: Example Partitioning of Functional Groups

HullDiscipline

Machinery andUtilities

CargoHandling

ElectricalFunctional

Group

ManeuveringFunctional

Group

PropulsionFunctional

Group

Vessel ServiceFunctional

Group

Navigation &Communications

FunctionalGroup

Line and PropellerShafting, Shaft

Support Bearings

ReductionGear

DieselEngine Propeller

Engine Support SystemsLube Oil, Water, Fuel, Hydraulic, Air,

Exhaust, Control Systems, Monitoring...

BasicEngine

Cylinder Linerand CylinderLubricationAssembly

Piston with Rodand Stuffing Box

Assembly

CylinderCover

Assembly

Crosshead withConnecting Rod

Crankshaft,Thrust Bearing

and TurningGear Assembly

FrameCrankcaseAssembly

ExhaustValve

MechnicalControl Gear,

Chain Drive andCamshaft

VibrationCompensatorsand Dampers

EquipmentItems

Subsystems

Systems

FunctionalGroups



Figure 3: Example System Block Diagram

Control Systems

Fuel System

BridgeSignal Governor

Bridge Signal

SpeedControlSignal

Heavy Fuel Oil

Diesel Oil

Air Starting System

Start/Stop Signals

BarringInterlockSignal

TorqueCrankcase

Vapor System

ControlAir

StartingAir

Vapor LubeOIl

Oil Sludge to Sludge Tank

Clean Vapor

Torque &Vibration

ContaminatedLube Oil

Cleaning System,Stuffing Box Drain Oil

To PropulsionShafting

To Lube OilSump Tank

Sludge(to waste)

LubeOIl

LubeOIl

ExhaustGasses& Noise

Scavenge Air

Cool Lube Oil

Lube Oil & Heat

Camshaft LubeOil System

Main LubeOil System

CylinderLubricating Oil

System

PressurizedFuel

Instrumentation& Alarms

RPM, Pressure,Temperature, Level

Cylinder Lubricating Oil

EngineRPM

Lube Oil

Lube Oil &Heat

LubeOIl

Lube OIl& Heat

Central Cooling Water System

CoolFreshwater

Freshwater& Heat

CoolFreshwater

Freshwater& Heat

Coo

l Fre

shw

ater

Fres

hwat

er &

Hea

t

Fres

hwat

er&

Hea

t

Fres

hwat

er

Exh

aust

Gas

ses

& N

oise

Con

dens

ate

Atm

osph

eric

Air

Seawater Sewater &Heat

CylinderLubricating

OilAlarms Readouts

Basic Engine

Scavenge Air &Exhaust Gas Systems,

includingTurbochargers



TABLE 3: Example Consequence/Severity Level Definition Format

Severity Level

Example Descriptors for Severity Level

Directional Control,

Propulsion, etc.

Explosion/Fire Loss of Containment Safety (1)

1 Minor, Negligible

Function is not affected, no significant

operational delays. Nuisance.

No damage to affected equipment or compartment, no

significant operational delays.

Little or no response necessary

Minor impact on personnel/No impact

on public

2 Major,

Marginal, Moderate

Function is not affected, however,

failure detection/corrective

measures not functional. OR

Function is reduced, resulting in

operational delays.

Affected equipment is damaged,

operational delays

Limited response of short duration

Professional medical treatment for

personnel/No impact on public

3

Critical, Hazardous,

Major, Significant

Function is reduced, or damaged machinery, significant

operational delays

An occurrence adversely affecting

the vessel’s seaworthiness or

fitness for service or route

Serious/significant commitment of resources and

personnel

Serious injury to personnel/Limited impact on public

4 Catastrophic, Critical

Complete loss of function

Loss of vessel or results in total

constructive loss

Complete loss of containment. Full scale response of

extended duration to mitigate effects on

environment.

Fatalities to personnel/Serious impact on public

Notes: 1 Safety losses are not intended to be compared to other losses to determine monetary equivalency.

TABLE 4: Probability of Failure (i.e., Frequency, Likelihood) Criteria Example Format

Likelihood Descriptor Description

Improbable Fewer than 0.001 events or < 1 event per 1000 vessels per year

Remote 0.001 to 0.01 events or 1 event per 100 to 1000 vessels per year

Occasional 0.01 to 0.1 events or 1 event per 10 to 100 vessels per year

Probable 0.1 to 1 events or 1 event per 1 to 10 vessels per year

Frequent 1 or more events or >1 event per vessel per year



Figure 4: RCM Task Selection Flow Diagram

A Select a failure mode for evaluation

B Select a cause for evaluation

Is the failure mode risk inthe highest or lowest risk

categories?

Is there high confidencein the failure mode

risk ranking?

One-time changerequired

Specify run-to-failure strategy C

Is there a condition-monitoring task thatis applicable and effective?

Specify condition-monitoring task at ½

the P-F interval

Considerredesign

Specifyone-timechange

Does the cause exhibit a wear-in and/orwear-out failure characteristic?

Is there a one-time change that isapplicable and effective?

Is there a planned maintenance task that isapplicable and effective?

Specify plannedmaintenance at the

appropriate life limit

Specify combination tasks at½ the P-F interval and

the life limit

Is there a combination of condition-monitoring and planned maintenance

tasks that are applicable and effective?

Will the loss of function from this causebe hidden or evident?

Is there a task(s) that is applicable andeffective?

Is there a failure finding task that isapplicable and effective?

Specify the tasksat the appropriateinterval to achieve

a tolerable risk

One-time changemay be necessary

to achieve atolerable risk

CSpecify failure-findingtask at the appropriate

interval

Yes

No

Yes Highest

LowestNo

Yes

No

Yes

Wear-out

Wear-in

Yes

No

Yes

Yes

No

No

Hidden

Evident

No

Yes

NoYes

No



Figure 4 (continued): RCM Task Selection Flow Diagram

C

Is there another cause associated with thisfailure mode to be evaluated?

Reevaluate the risk assuming the selectedmaintenance tasks and any one-time

changes are in place.

Does the risk level meet the riskacceptance criteria?

Is the risk level tolerable and no further riskreduction is practically feasible?

Reevaluate the maintenance tasks andone-time changes for the failure mode.

B

Evaluate the next failure mode.

A

A

Yes

No

No

No

Yes

Yes

Step No. 7 – Spare Parts Holdings An additional feature of the ABS RCM Program is a requirement for the selection of spare parts applying risk principles. We adapted from Figure 14.1 of NES 45 (Ministry of Defence (UK) 1999) the diagram shown in Figure 5. We have shown the spare parts decision diagram with an example of determination of the spare parts for a fuel oil supply pump. As with the FMECA, the operating context of the equipment is an important factor in determining spare parts holdings. In the example, the two supply pumps are operated alternately on a weekly basis so that after a period of time, both pumps will have roughly the same number of operating hours. The example indicates holding a bearing replacement kit onboard can reduce the risk associated with the vessel being out of service because of two inoperable pumps.

If the operating context was changed to running one pump until maintenance was required, then operating the spare pump until it is shut down for maintenance, a different conclusion could be reached. In this case, since the duty pump will have more operating hours than the standby pump, the vessel operator would have cause to believe the frequency of the standby pump being inoperable at the time the duty pump is shut down to be much lower than in the example above. The higher availability of the standby pump could be confirmed by satisfactory failure finding tests over a “long” period of time. For this case, provided spare parts can be delivered to the vessel “quickly”, ordering spare parts instead of holding them onboard may be an acceptable spare parts strategy.



Figure 5: Example of Use of Spares Holding Decision Flow Diagram

W ill the stock-ou t o r the stock-out and fu rther fa ilure have aneffect on any o f the fo llow ing end effects?

P ropulsion , D irectional C ontro l, L oss of C ontainm ent,E xp losion /F ire , D rilling , P osition M ooring , H ydrocarbonP roduction and Processing , Im port/E xport F unctions

F ailu re of the standby pum p in the event of fa ilure o f theoperating pum p w ill cause loss of p ropulsion . A nsw er Y E S .

N o spareshold ingrequ ired

1. C an the parts requ irem ents be antic ipated (e .g ., can theparts be obtained before failure is expected to occur)?

D egradation of pum p bearings can be m onitored w ithcondition m onito ring program presen tly im plem ented .Parts can be ordered/delivered w ith in 7 days. Y E S .

2 . D oes the stra tegy , o rdering parts befo re dem and occurs,provide an accep tab le risk?

In the even t o f standby pum p failu re , vesse l w ill be ou t ofserv ice as long as 7 days. R isk is unaccep tab le . N O .

O rder partsbefore

dem and

1 . Is it feasib le and cost-effective to ho ld required parts andquan tity in sto res?

B earing rep lacem ent k it costs $X . S ize and w eigh t areinsignificant. V essel c rew is qualified to repair pum p.Y E S .

2 . D oes the stra tegy , ho ld parts, p rov ide an acceptab le risk?

In the even t of standby pum p fa ilure , repairs can becom pleted in 4 hours. Y E S .

H old bearingrep lacem ent k it

onboard

R evise/R eviewR C M T asks

Y es

N o

Y es

N o

N o

Y es

Example Operating Context and Analysis. A Fuel Oil piping system is provided with two fuel oil supply pumps arranged in parallel redundancy. Each pump is sized so as to supply heavy fuel oil to the main propulsion engine and two of the three diesel generator engines operating at their maximum continuous rating. The pumps are operated as follows: the No. 1 pump is operated for one week at a time with the No. 2 pump on standby. After one week, the No. 1 pump is secured and put on standby and the No. 2 pump is operated for one week. Anticipated annual service hours for both pumps are the same.

Step No. 8 – Sustainment Process

Any successful maintenance program needs to be dynamic to address modifications to systems and their respective equipment and effects of aging for the life of the machinery. A process for providing feedback is necessary and is referred to as RCM sustainment.

The objective of the sustainment process is to: • Continually monitor and optimize the current

maintenance program • Delete unnecessary requirements • Identify adverse failure trends



• Improve overall efficiency and effectiveness of the RCM and maintenance programs

ABS has listed several sustainment tools in the RCM Guide as an aid to the vessel owner/operator when conducting the sustainment process. These are:

• Trend analysis • Maintenance requirements document reviews • Task packaging reviews • Age exploration tasks • Failures • Relative ranking analysis • Other activities For example, in the case of unexpected machinery

failure, ABS would recommend use of the Failures tool based on Figure 5-5 of Naval Air Systems Command (USA), Guidelines for the Naval Aviation Reliability-centered Maintenance Process, NAVAIR 00-25-403 (Naval Air Systems Command (USA) 2001). This process is illustrated in Figure 6. A root cause analysis is performed first to develop an understanding of the failure and includes these steps:

• Identifying the failure or potential failure • Classifying the event and convening a trained

team suitable for addressing the issues posed by this event

• Gathering data to understand how the event happened

• Performing a root cause failure analysis to understand why it happened

• Generating corrective actions to keep it (and similar events) from recurring

• Verifying that corrective actions are implemented

• Putting all of the data related to this event into an information system for trending purposes

The failure may be addressed by corrective actions for which an RCM analysis is not necessary. Examples of non-RCM corrective actions include technical publication changes and design changes.

The root cause analysis may reveal problems that may need immediate attention. Issuing inspection bulletins, applying temporary operational restrictions

and implementing operating safety measures are examples of interim actions.

The results produced from reviewing the RCM analysis will be a factor that should be considered in determining a response to the failure. It is necessary that an RCM review be part of the overall methodology. The RCM review and update, if required, will determine if changes in maintenance requirements are necessary. The review will indirectly aid in determining if corrective actions are necessary. Decisions not to update the RCM analysis should be documented for audit purposes. During the RCM review, the following questions should be addressed:

• Is the failure mode already covered? • Are the failure consequences correct? • Are the reliability data accurate? • Is the existing task (or requirement for no task)

adequate? • Are the related costs accurate?

When new failure modes or failure modes previously thought unlikely to occur are determined to be significant, the RCM analysis is to be updated. The existing analysis for a failure mode may also be determined to be correct or inadequate. Inadequate analyses can result for any number of reasons, such as revision of mission requirements, changes to operating context or changes to maintenance procedures.

Failures and other unpredicted events are available from several sources, including the following examples:

• Defect reports issued by maintenance engineering or the vessel’s crew

• Defects discovered during routine vessel repairs in a shipyard

• Vendor and original equipment manufacturer reports related to inspections, rework or overhauls

• Design changes, which may be in the form of a single item change or a major system modification

• Results of tests (such as certification tests or tests performed during the course of a failure investigation or some other unrelated event) that may require RCM review and update



Figure 6: Process to Address Failures and Unpredicted Events

Failure Root CauseFailure Analysis

Non-RCMCorrective Action

Required?

Non-RCMCorrective Action

Interim ActionRequired? Interim Action

RCM Review

RCM UpdateRequired? RCM Update

Document Results

No

Yes

No

Yes

Yes

No

CONCLUSIONS We have described some of the processes used in the RCM Guide such as risk assessment techniques, failure management task selection and sustainment. We have also described some of the approaches we have taken to address issues such as consistency among analyses received from different owner/operators, lack of quantitative data and identifying consequences objectively. RCM is a relatively new maintenance approach in the marine industry and time will be needed before the industry becomes familiar with the processes. The RCM Guide provides a thorough and sound basis while maintaining a practical approach to current marine maintenance practices.

We believe over time as the marine industry becomes familiar with the application of RCM techniques, vessel owners/operators will see the same benefits of other industries that have embraced RCM. Some of the benefits that vessel owners/operators can expect are:

• An integrated program to address safety and environmental concerns;

• Increased integrity and reliability of critical machinery and components;

• More cost-effective maintenance; and • Improved understanding of equipment failures

and their impact on vessel performance.



ACKNOWLEDGEMENTS The development of the RCM Guide and RCM Guidance Notes has been a team effort and would not have been possible without significant contributions from Ah Kuan Seah, Yoshi Ozaki, Randal Montgomery and Piyush Parikh. Special thanks are given to the members of the specially created RCM Committee for their thoughts and comments and in particular, Kenneth Gardner for his comments during system template development.

REFERENCES American Bureau of Shipping, 2004, “Guidance Notes for Reliability-centered Maintenance,” July 2004.American Bureau of Shipping, 2003, “Guide for Surveys Based On Reliability-centered Maintenance,” December 2003.International Maritime Organisation, 2001, International Code of Safety for High-Speed Craft, 2000, Annex 3, Use of probability concept, Annex 4, Procedures for failure mode and effects analysis. Ministry of Defence (UK), Requirements for the Application of Reliability-centered Maintenance to HM Ships, Submarines, Royal Fleet Auxiliaries, and Other Naval Auxiliary Vessels, (Naval Engineering Standard NES 45, Issue 3), Bath, UK, September 1999. Moubray, John, “Reliability-centered Maintenance-2nd edition”, Industrial Press Inc., New York, 1997. Naval Air Systems Command (USA), Guidelines for the Naval Aviation Reliability-centered Maintenance Process, (NAVAIR 00-25-403), 01 February 2001. Nowlan, F.S. and Heap, H.F., “Reliability-Centered Maintenance”, U.S. Department of Commerce, National Technical Information Service, 1978. Smith, Anthony M., “Reliability-Centered Maintenance,” McGraw-Hill, New York, 1993. Society of Automotive Engineers, Evaluation Criteria for Reliability-Centered Maintenance (RCM) Processes (SAE JA1011), Warrendale, PA, 1999.

APPENDIX 1 – ANALYSIS OF RISK REDUCTION The benefits of employing the suggested maintenance tasks and one-time changes can be seen in the anticipated reduction in risk. Table 5 is a summary of

the risk associated with an example analysis for loss of propulsion. The number in the matrix cells indicates the number of loss events with the cell’s corresponding frequency and severity. CR indicates the number of Current Risk events, (e.g. with the current preventative maintenance plan) and PR indicates the number of Projected Risk events (e.g. with the proposed preventative maintenance plan). The data in Table 5 are obtained from a Maintenance Task Selection Worksheet, an example format is shown in Table 6. The Current Risk and the Projected Risk for each failure mode is tabulated. For cases where several tasks are selected for a failure mode, the task with the highest risk is used in the tables. From the example analysis, it can be seen from a qualitative aspect that there will be a reduction in overall risk by applying the selected maintenance tasks. The risk reduction from a quantitative aspect can best be estimated in the following manner. Table 7 provides the current frequency, projected frequency and frequency reduction for the loss evaluated. Because the frequency categories are ranges, ranges with an upper and lower bound represent the frequencies. This table is developed from the data presented in Table 5. To determine the frequency reduction for Severity Level 4 for Propulsion in Table 7, we refer to the Severity Level 4 row in Table 5. To calculate the Current Events/yr upper bound in Table 7, Severity Level 4, we note there is one Current Risk in the “Remote” column and three Current Risks in the “Improbable” column of Table 5. From Table 4, the frequency range for Remote is 0.001 to 0.01 events/yr and for Improbable, <0.001 events/yr. The Current Events/yr upper bound is 1 * (0.01) + 3 * (0.001) = 0.013 and Current Events/yr lower bound is 1 * (0.001) + 3 * (0.000) = 0.001. The Projected Events/yr is calculated similarly. The Frequency reduction is determined by subtracting the Projected Events/yr from the Current Events/yr for the upper bound and for the lower bound. For Severity Level 4, the proposed maintenance tasks are projected to reduce the frequency of a Severity Level 4 event by 0.001 to 0.009 events/yr. If an economic value is assigned to the Severity Level, for example, >$1,000,000, an annual economic risk reduction can be estimated: (0.001) * $1,000,000 to (0.009) * $1,000,000 = $1,000 to $9,000 per year risk reduction. Similar calculations are done for the other severity levels.

TABLE 5: Propulsion Category Risk Matrix

Likelihood of Failure Severity Level

Improbable Remote Occasional Probable Frequent 4 PR – 4, CR – 3 CR – 1 3 PR – 10, CR – 6 PR – 5, CR – 6 CR – 3 2 PR – 2 CR – 2 1 PR – 2 CR – 2



Risk Shade High

Medium Low

TABLE 6: Example Maintenance Task Selection Worksheet

No.: Description:

Effects Risk Characterization Item Failure Mode

Failure Char.

H/E

Local Functional failure

End S CL CR

No.: Description:

Task Selection Item

Proposed Action(s)

PL PR Disposition

Symbol Description Failure Characteristic Enter failure description such as wear-in , random or wear-out

failure or combination H/E Hidden failure/evident failure

S Severity level CL Current likelihood (frequency) of failure CR Current risk

Proposed Action(s) Proposed Maintenance for failure mode PL Proposed likelihood of failure applying Proposed Maintenance PR Risk after applying proposed maintenance

Disposition Note as to whether proposed maintenance will be applied

TABLE 7: Expected Event Frequencies for Propulsion

Severity Categories Severity Level 1 Severity Level 2 Severity Level 3 Severity Level 4

Upper Bound 0.02 0.2 0.366 0.013

Current Events/yr Lower

Bound 0.002 0.02 0.036 0.001

Upper Bound 0.002 0.02 0.06 0.004

Projected Events/yr Lower

Bound 0 0.002 0.005 0.0

Upper Bound 0.018 0.18 0.306 0.009 Frequency

Reduction Events/yr Lower

Bound 0.002 0.018 0.031 0.001


Documents

Development of Machinery Survey Requirements Based on Reliability